Host laboratory: Fire Experimental Laboratory (LEF)
Beggining of the thesis: February 2012
Student name: Arnaud Brunner
Description of the subject
A fire that breaks out in a nuclear facility can lead to under-ventilated fires. An increase in the overall production of soot generally ensues on account of the vitiation of the surrounding environment. This in return modifies the retroaction of the radiation of the flame towards the combustible material and thus influences the pyrolysis process. It is consequently necessary to understand the effects on combustion parameters of the level of oxygen surrounding the fire. The main parameters that influence combustion are the soots and gases stemming from the combustion, the rate of pyrolysis of the combustible material, the radiative transfer of the flame towards its environment and in return on the combustible material, as well as the temperature and the geometry of the flame.
Most of the models currently used in calculation codes rely on experimental data obtained in the main from tests carried out in laboratory cone calorimeters. The pyrolysis mechanisms are governed by heat transfer by radiation and by convection at the surface of the combustible material. The dominant mechanisms are convective transfer for combustible materials of small dimensions and radiative transfer for those of large dimensions. For pool fires, the transition between the two regimes takes place around a diameter of 0.2 m. For the dimensions tested on laboratory calorimeters, the regime is mainly convective and an external input of radiation is necessary in order to get closer to the real situations of a fire.
In order to exploit these results on a real scale, it is necessary to evaluate the influence of scale effects. The most conventional approach for dealing with this aspect is to carry out tests for the different characteristic dimensions of the combustible material and thus to determine a behavioural law. This approach, applied to pool fires in open environments, is more difficult to implement for studies under controlled atmosphere due to the difficulties of maintaining a uniform distribution of the oxidant around the centre of the fire and the difficulties of carrying out tests on a larger scale.
The objective is to study the effects of under-ventilation on the combustion parameters in dominant radiative regime. Under-ventilation consists in reducing the oxygen content of the oxidant. The dominant radiative regime approach consists in carrying out tests with fire centres of characteristic dimensions greater than 0.2 m.
We propose carrying out representative tests of real fires in a confined calorimeter environment under controlled atmosphere on a medium scale. The test apparatus set up within IRSN will make it possible to build up a unique experimental data base on combustion parameters. The combustion products emitted from the device (CO, CO2, unburned material) will be analysed, the concentration and the morphologies of the soot will be measured and an evaluation of the radiative transfer of the flame will be carried out and its overall geometry will be studied. The techniques for measuring the combustion parameters of interest either already exist or need to be developed within the context of the PhD work.
After a bibliographic study, the work will consist in an experimental and metrological qualification of the calorimeter. In order to be free of pyrolysis, a first series of tests will be carried out on gaseous combustible materials. These flame-free measurements under stationary conditions will enable, in the first instance, the limits of validity of existing models on combustion parameters in under-ventilated environments to be determined. The work will then be extended to include other series of tests conducted using liquid and solid combustible materials. These experimental results will in parallel serve for the development and the validation of combustion and radiation models.